答辩公告
我的位置在: 首页 > 答辩公告 > 正文
NKUNZIMANA Hilaire答辩公告
浏览次数:日期:2018-10-19编辑:研究生教务办1

答辩公告

论文题目

Computational Resource   Scheduling Algorithm under Quality of Service Constraints in Cloud Computing

答辩人

NKUNZIMANA Hilaire

指导教师

李肯立 教授

答辩委员会

主席

陈志刚 教授

学科专业

计算机科学与技术

学院

信息科学与工程学院

答辩地点

国家超级计算长沙中心二楼会议室

答辩时间

20181101

下午15:30


学位论文简介

As a new type of business computing models, cloud computing uses virtualization technology to connect a large number of computing resources, storage resources, and software resources to construct a large-scale shared virtual resource pool, providing customers with cheap and on-demand services. At the same time, users in various application fields usually focus on the high-quality user experience of cloud computing, such as Quality-of-Service (QoS), Service Level Agreement (SLA), and security and privacy requirements. Computational resource scheduling is an important part of cloud computing. The QoS level directly affects the quality and utilization of users' cloud computing service. Moreover, the overall consumption and performance of all computing resources have a strong impact on the interests of service providers. The quality of service of cloud computing needs to be guaranteed through the computational resource scheduling and task scheduling, which raise new demands and challenges in cloud computing resource scheduling. This thesis focuses on the computational resource scheduling in cloud computing environments under the QoS constraints. The main work and innovations of this thesis are as follows:

(1) We introduce the hardware and software architecture models of the cloud computing environment considered in this thesis and describe the components and characteristics of the architecture model in detail. Then, the overall QoS constraint model of cloud computing is proposed from the aspects of performance, operational reliability, data security, and cost. Finally, measurable and reusable measurement metrics are defined to evaluate the QoS constraint model. We demonstrate how the proposed QoS metrics can be used to provide more accurate and scientific evaluation metrics for computational resource scheduling.

(2) We propose a cloud computational resource scheduling algorithm based on a genetic algorithm. First, the genetic algorithm is improved and applied to solve the problem of virtual machine allocation in cloud computing. Considering that the solution to the problem is the result of multi-objective optimization, we satisfy each QoS constraint by calculating the finite value of each chromosome, so that the quality of the solution meets the user's actual application requirements. We describe the modeling process for chromosomes and genes that correspond to the virtual machine scheduling problems.

(3) We propose an enhanced cloud computing deployment model to facilitate the application of cloud computing technology in developing countries, and solve practical problems involved in the deployment of cloud computing environments. In developing countries, due to the lack of a stable power supply, service availability of cloud computing resources becomes another important problem in addition to energy consumption. Firstly, we describe the business management processes of cloud computing services in developing countries, and identify all security threats and potential risks that compromise the cloud computing services. Then, we propose an enhanced cloud computing deployment model that can effectively fault-tolerant in a power-off state and provide continuous service.

主要学术成果

[1] Hilaire NKUNZIMANA}, Kenli Li. Enhanced cloud computing development model to address security concerns in developing countries context [J]. \textit{American SCI Journal}, 2018, 435: 124-149. (SCI, 第一作者)